Earth and Planetary Science Letters 238 (2005) 156 – 171

www.elsevier.com/locate/epsl

Late Archean to Early Paleoproterozoic global tectonics,

environmental change and the rise of atmospheric oxygen
Mark E. Barley a,*, Andrey Bekker b,1, Bryan Krapež a,2
a
School of Earth and Geographical Sciences, The University of Western Australia, 35 Stirling Highway, Crawley WA 6009, Australia
b
Geophysical Laboratory, Carnegie Institution of Washington, 5251 Broad Branch Road, Washington DC 20015, USA
Received 23 August 2004; received in revised form 14 June 2005; accepted 24 June 2005
Available online 24 August 2005
Editor: E. Boyle

Abstract

Analysis of the tectonostratigraphic records of Late Archean to Early Paleoproterozoic terranes indicates linkage between
global tectonics, changing sea levels and environmental conditions. A Late Archean tectonic cycle started at ~2.78 Ga
involving the breakup of a pre-existing continent (Vaalbara) and the most prodigious period of generation and preservation of
juvenile continental crust recorded in Earth history during a period of plume breakout (~2.72 to 2.65 Ga) accompanied by high
sea levels. During this period, cratons formed by accretion of granitoid–greenstone terranes at convergent margins started to
aggregate into larger continents (e.g. Kenorland). Lower sea levels between ~2.65 and 2.55 Ga were followed by a second
(~2.51 to 2.45 Ga) period of plume breakout resulting in a global peak in magmatism, high sea levels and deposition of banded
iron formations (BIF) on the trailing margins of the Pilbara and Kaapvaal cratons. Cratons in South Australia, Antarctica, India,
and China record convergent margin magmatism, orogeny and high-grade metamorphism between 2.56 and 2.42 Ga.
Continued aggregation of continental fragments (e.g. amalgamation of Indian cratons) may have formed the Earth’s first
supercontinent by ~2.4 Ga with a return to low sea levels and relative tectonic quiescence before the supercontinent started to
breakup from ~2.32 Ga.
Although oxygenic photosynthesis had evolved by 2.71 Ga, the irreversible rise of atmospheric O2 to N 10 5 PAL appears to
have occurred between 2.47 and 2.40 Ga following the second plume breakout and coinciding with a decline in BIF deposition
and the maximum extent of the supercontinent suggesting dynamic linkage between tectonics and both the sources and sinks of
oxygen. Periods of plume breakout (2.72 to 2.65 Ga and 2.51 to 2.45 Ga) would have limited ocean productivity and the rate of
photosynthesis and also enhanced the reduced conditions typical of the Archean biosphere, as well as the greenhouse gas
contents of the atmosphere necessary to maintain temperate conditions. This suggests that either an increase in the oxidation
state of volcanic gasses during the second plume breakout, or a decreased flux of reduced gasses following plume breakout,
coupled with the filling of crustal oxygen sinks and possibly also an increase in ocean productivity and the rate of
photosynthesis resulted in the global flux of reduced gasses falling below oxygen production leading to a rise of atmospheric

* Corresponding author. Tel.: +61 8 64887322; fax: +61 8 64881037.

E-mail addresses: [email protected] (M.E. Barley), [email protected] (A. Bekker), [email protected] (B. Krapež).
1
Tel.: +1 202 4787974; fax: +1 202 4788901.
2
Tel.: +61 8 64882771; fax: +61 8 64881037.

0012-821X/$ - see front matter D 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.epsl.2005.06.062
 M.E. Barley et al. / Earth and Planetary Science Letters 238 (2005) 156–171 157

O2 accompanied by loss of the CH4-rich greenhouse atmosphere resulting in the Earth’s first widespread glaciation. Detrital
pyrite and uraninite in 2.45 to 2.40 Ga sediments suggests that terrestrial surface environments were not yet extensively
oxidized. The oldest evidence of extensive oxidative weathering is associated with 2.32 to 2.22 Ga glacial deposits and breakup
of the supercontinent.
D 2005 Elsevier B.V. All rights reserved.

Keywords: Archean; Paleoproterozoic; global tectonics; environmental change; oxygen

1. Introduction biomarkers in sedimentary rocks as old as 2.71 Ga

[11], and may have been associated with much older
The Late Archean to Early Paleoproterozoic (2.8 to microfossils and stromatolites. Furthermore, the fact
2.2 Ga) is a key period in Earth history. It records one that d 13C values of carbonates have remained near
of the most important episodes of growth and preser- 0x throughout most of Earth history suggests that
vation of continental crust followed by the transition organic carbon has been buried at a rate more-or-less
from oxygen poor to oxygenated conditions in the proportional to its production and input to the atmo-
atmosphere and widespread glaciation. The recent sphere [7]. If so, sinks for oxygen rather than its
discovery of non-mass dependent fractionation source may have changed suggesting a geological
(NMDF) in sulfur isotopes has provided a new tool rather than biological control on the timing of the
for tracing changes in the oxygen content of the atmo- rise in O2. Kasting et al. [12], Kump et al. [7], and
sphere and reduced the uncertainty regarding the tim- Holland [8] have argued that changing redox state of
ing of the initial rise in atmospheric oxygen [1–3]. volcanic gasses resulted in irreversible oxidation of
Sulfur from Archean sedimentary sulfides and sulfates the atmosphere at some time between 2.5 and 2.2 Ga.
has D33S values that range from 2.5 to + 8.1x In contrast Catling et al. [6] and others have argued
whereas sulfides and sulfates younger than ~2.32 Ga that the oxidation state of volcanic gasses has
have D33S signals b0.4x [3]. Preservation of large remained approximately constant and that higher
NMDF signals in the Archean requires an anoxic Archean atmospheric CH4 contents resulted in faster
atmosphere; lack of NMDF in rocks younger than escape of H2 to outer space. In this scenario, irrever-
2.32 Ga is thus consistent with significant quantities sible oxidation of the atmosphere would have
of oxygen (N 10 5 PAL, present atmospheric levels) in occurred when available crustal oxygen sinks were
Earth’s atmosphere. The youngest well dated sedi- satisfied such that the global flux of reduced gasses
mentary sulphides with a NMDF signal N 0.4x are fell below the photosynthetic rate of oxygen produc-
hosted by 2.47 Ga banded iron formations (BIF) of tion [6,13]. Bjerrum and Canfield [9] have also
the Brockman Supersequence in the Pilbara Craton pointed out that adsorption of phosphorous onto iron
[1,4]. The oldest sedimentary sulphides with a strati- oxides deposited as BIF during the Late Archean and
graphically persistent NMDF signal that is b 0.4x Paleoproterozoic would have inhibited organic pro-
occur in shales of the Lower Huronian Supergroup ductivity reducing rates of photosynthesis and contri-
in the Superior Craton that directly underlie the first buting to low concentrations of atmospheric oxygen.
Huronian glacial diamictite [5]. If, as we argue below, Resolution of the cause of the rise in O2 requires a
the oldest glacial diamictites in the Superior and much better understanding of the balance between
Pilbara cratons can be correlated, it is most likely oxygen sources and sinks including the dynamic
that the transition from an anoxic atmosphere to an role of those sinks and element cycles that are con-
atmosphere with oxygen N 10 5 PAL occurred trolled by global tectonics (cf. [10]).
between 2.47 and 2.40 Ga. Links between Phanerozoic tectonics and environ-
Although the timing of the rise of atmospheric O2 mental change are well documented, with greenhouse
is becoming better-established, its cause remains hotly conditions and high sea level accompanying periods
debated [6–10]. Cyanobacteria, the first oxygenic of enhanced global magmatism (plume breakout or
photosynthesizers, have been identified from organic superplume events) during the breakup and aggrega-
158 M.E. Barley et al. / Earth and Planetary Science Letters 238 (2005) 156–171

tion of supercontinents followed by periods of low lish correlation of key events and environmental indi-
sea-level and episodic glaciation (Fig. 1). For plausi- cators. The picture that emerges is of a Late Archean
ble parameters computer modeling of mantle layering to Early Paleoproterozoic tectonic cycle (including the
and convection through geological time [14] predict largest plume breakout event preserved in the rock
episodic mantle overturn and plume breakout during record), linked to changing relative sea levels culmi-
the Archean (replacement of upper mantle by deeper nating in formation of the Earth’s first supercontinent
hotter mantle) with a major event at ~2.7 Ga possibly by 2.4 Ga. Oxidation of the atmosphere and loss of
involving orders of magnitude greater heat flux and Late Archean greenhouse conditions resulting in the
mafic magmatism than Phanerozoic plume breakout Earth’s first widespread glaciations followed soon
events. Although it is widely accepted that plume after a second period of plume breakout coinciding
breakouts accompanied formation of a supercontinent with a decline in BIF deposition during the final
or supercontinents during the Late Archean to Early stages of amalgamation of this supercontinent sug-
Paleoproterozoic [15–20], links between global tec- gesting dynamic linkage between the source and
tonics and environmental conditions during this per- sinks of oxygen and global tectonics. Breakup of the
iod are not well understood. As a result, the transition continent was accompanied by episodic glaciation and
from the Late Archean to the Early Paleoproterozoic oxidative weathering of the continental crust.
is generally viewed as a secular change from a tec-
tonic regime dominated by mantle plume activity to a
quiet period with little evidence of modern tectonic 2. Phanerozoic tectonics and environmental change
processes.
We here review the tectonostratigraphic records of Although the Earth’s tectonic regime may have
several cratons using precise geochronology to estab- changed since the Archean, links between Phanero-

Phanerozoic Tectonic Cycles

Breakup Pangea Breakup
Pannotia Assembly Pangea
200
Relative Sea Level (m)

PBE PBE
100

PBE
Sea Level

*
G *G * G

600 500 400 300 200 100 0

Age (Ma)

NP C Ord S Dev Carb P Tr Jur Cret Cz

Fig. 1. Phanerozoic tectonic cycles and eustatic sea level relative to present sea level over the past 600 million years. PBE = plume breakout
event, G = glaciation.
 M.E. Barley et al. / Earth and Planetary Science Letters 238 (2005) 156–171 159

zoic tectonics and environmental change provide a combined effects of plume breakout events and global
useful background for this analysis of the Archean tectonics.
to Paleoproterozoic rock record. Global tectonics can Because a range of tectonic processes are occurring
be viewed as the cyclic breakup, dispersal and assem- in different places on the Earth at any one time, links
bly of supercontinents or the cyclic contraction and between global tectonics and environmental condi-
expansion of the external ocean (currently the com- tions are complex with changes reflecting the effects
bined Pacific and Indian Oceans). Because the move- of those processes which have a dominant input at a
ment of fragments of continental crust is controlled by particular time. Aggregation of fragments of continen-
the production and subduction of ocean crust at plate tal crust to form a supercontinent tends to favor global
margins, it is unlikely that each global tectonic cycle low sea levels and cooler climates [32,33]. This is
involves the complete breakup and re-aggregation of a because rapid uplift and exposure along sutures during
supercontinent, rather that periods when continental the assembly of continental fragments will result in
dispersal is the dominant tectonic process are fol- enhanced erosion and chemical weathering which
lowed by periods when continental assembly is domi- together with biological productivity draw CO2
nant forming new supercontinents that include most, down from the atmosphere. This combined with
but not necessarily all, continental fragments. The increasing albedo (linked to land–water ratio) can
rifting and breakup of continents forms internal or lead to large continental ice sheets. However, tectonic
Atlantic-type oceans which may either close or stasis during the maximum extent of a supercontinent
become part of the external ocean as the next super- will depress both physical erosion and chemical
continent forms with these tectonic events recorded by weathering allowing atmospheric CO2 levels to rise
the rock records of continental margins and orogens as and resulting in warmer climates. Breakup of the
described by Murphy and Nance [21,22] and Krapez supercontinent, with uplift during the early stages of
[23]. Phanerozoic tectonic evolution can be described rifting, will initially favor erosion and weathering
in terms of one and a half ~360 Ma cycles of super- thereby drawing down atmospheric CO2 and resulting
continent breakup and aggregation [15,23–25]. Dur- in cool climates. Enhanced volcanism during breakup
ing the Late Neoproterozoic (from ~590 Ma) and will result in new hydrothermally active spreading
Early Paleozoic, a supercontinent broke up. Continen- ridges and subduction zone magmatism, elevating
tal plates later reassembled at the end of the Paleozoic sea levels, and promoting anoxia in the deep ocean.
to form Triassic Pangea, which broke up during the Elevated CO2 levels, due to mantle degassing, pro-
Mesozoic with continental separation closely linked to mote warmer climates and increased weathering rates
mantle plume magmatism [26]. with increased deposition of carbonate, both as sedi-
Although the present distribution of mantle plumes ment on continental margins and by submarine hydro-
appears independent of plate-scale mantle flow, peri- thermal alteration. Expansion of anoxic water onto
ods of constant magnetic field polarity lasting up to 40 continental shelves and into intracontinental basins
Ma and enhanced global magmatism during the Ordo- increases burial of organic carbon. The environmental
vician (~500–460 Ma), Permo–Carboniferous (~300– effects of plume breakout (high sea level, ocean
260 Ma), and Cretaceous (120–80 Ma) are interpreted anoxia, elevated CO2, increased carbonate and
to represent elevated mantle plume activity (Fig. 1), organic carbon burial, warm climates) will enhance
referred to as plume breakout or superplume events the effects of continental breakup (possibly to the
[20,27,28]. While numerous large igneous provinces extent of nullifying the effects of initial uplift and
provide strong evidence for periods of plume breakout weathering of rifting continental margins), and may
during the Cretaceous and Late Archean to Protero- also either nullify or moderate the environmental
zoic [20,29], evidence for the Paleozoic events is effects of supercontinent formation.
more ambiguous. Worsley et al. [30] and Nance et The Neoproterozoic saw the growth of a super-
al. [15] have proposed models linking sea-level, bio- continent, episodic global glaciation and a rise in
geochemical cycles and climate changes to supercon- atmospheric oxygen to levels comparable to present
tinent cyclity since the Archean, whereas Krapez [23], atmospheric abundances [34]. This was followed by
Barley et al. [29], and Condie [20,31] described the continental breakup and dispersal during the Cam-
160 M.E. Barley et al. / Earth and Planetary Science Letters 238 (2005) 156–171

brian and Ordovician, with volcanism and sedimenta- provide the best record of the breakup and assembly
tion in marginal basin environments and thick succes- of ancient continents and supercontinents [23], and
sions of sulfide-rich black shales (Fig. 1). The the abundance of large igneous provinces including
Ordovician saw one of the global peaks in production komatiites and coeval accelerated growth of juvenile
of volcanogenic massive sulfide deposits, episodic continental crust provide evidence for periods of
ironstone deposition, a peak in ophiolite obduction plume breakout [20,29,31].
(closure of marginal basins and terrane accretion), The Pilbara and Kaapvaal cratons (Figs. 2 and 3)
convergent margin magmatism, the largest explosive functioned as stable continental lithosphere at 2.8 Ga
volcanic eruption preserved in the Phanerozoic rock and may have been part of the continent Vaalbara [14].
record, and high sea level during plume breakout In the Pilbara, bimodal tholeiitic volcanic rocks and
[28,35–38]. This was followed by sea level drop associated clastic sedimentary rocks of the Nullagine
and Late Ordovician–Silurian glaciation [33], with Supersequence were deposited in continental graben
sea levels reaching a minimum during the Silurian. between 2.77 and 2.73 Ga [39,40]. These are overlain
As more continental fragments assembled during the by the Mt Jope Supersequence, a large igneous pro-
late Paleozoic to form Pangea, ~50% of the world’s vince comprising tholeiitic to mildly alkaline flood
coal reserves were deposited during the Permo–Car- basalt, with local komatiitic basalts and interlayered
boniferous plume breakout accompanied by high sea- stromatolitic carbonates deposited in an intracontinen-
levels, episodic marine anoxia, excursions in C and Sr tal rift basin between 2.73 and 2.715 Ga [40]. In the
sea-water isotopic signatures [20,28] and elevated northern Pilbara, lava flows were subaerial, whereas in
atmospheric oxygen levels. Although northern hemi- the south submarine pillowed flows are typical. The
sphere climates were warm at this time the southern overlying Jeerinah Formation deposited after 2.71 Ga
continent Gondwanaland experienced Permian glacia- records a marine transgression. This unit contains
tion [32] with sea-level reaching a minimum at the N 200 m of sulfidic black shales, and is intruded by a
Permo–Triassic boundary. suite of dolerite sills. The distribution of volcanic and
The breakup of Pangea was initiated by litho- sedimentary facies is consistent with successful rifting
spheric extension and coincident plume magmatism of the now southern margin of the craton. In the eastern
during the Triassic and Jurassic with the development Pilbara, the Jeerinah Formation is overlain by the
of extensive marginal basins [26]. Plume breakout shallow-water stromatolitic ~2.63 Ga Carawine Dolo-
during the Cretaceous saw the closure of many of mite [41]. In the western and central Pilbara, there is
these basins due to terrane accretion, a peak in con- apparently an unconformity at the base of the ~2.63 Ga
vergent margin magmatism, high sea level, oceanic Roy Hill Shale of the Jeerinah Formation that coin-
anoxia, extensive deposition of black shales (includ- cides with the transition from rifting to passive margin
ing ~60% of world’s oil reserves), and episodic sedimentation [41] indicating breakup of a continent
deposition of ironstones [20,27]. This also coincided and formation of an internal ocean.
with high concentrations of atmospheric CO2, ele- The Kaapvaal Craton records a similar history with
vated global temperatures and extensive deposition 2.72 to 2.64 Ga continental komatiitic basalts, tholeii-
of marine carbonates. Following the separation of tic basalts and sedimentary rocks of the Ventersdorp
Australia and Antarctica and India–Eurasia collision, large igneous province overlain by transgressive black
global climates have become cooler with late Ceno- shales and carbonates of the ~2.64 Ga Schmidtsdrif
zoic ice ages [32,33]. Subgroup of the Griqualand West Basin and Buffalo
Springs/Wolkberg and Black Reef Quartzite of the
Transvaal Basin [42]. The 2.78 to 2.60 Ga histories
3. Archean to Paleoproterozoic tectonics — 2.8 to of the Pilbara and Kaapvaal cratons provide clear
2.41 Ga evidence for a Late Archean period of continental
breakup enhanced by mantle plume magmatism.
In the absence of reliable paleomagnetic con- In contrast, granitoid–greenstone terranes world-
straints, the geological records of Late Archean to wide record the formation of new continental crust
Paleoproterozoic continental margins (Figs. 2 and 3) at convergent margins between 2.8 and 2.6 Ga. The
 M.E. Barley et al. / Earth and Planetary Science Letters 238 (2005) 156–171 161

Fig. 2. Comparison of the tectonic histories of representative Late Archean rifted and convergent margins during the Late Archean to
Paleoproterozoic.
162 M.E. Barley et al. / Earth and Planetary Science Letters 238 (2005) 156–171

eastern Yilgarn Craton in Western Australia and the In the eastern Yilgarn Craton, early ~2.78 to 2.72
Abitibi Belt in the Canadian Superior Province (Figs. Ga arc and back-arc basin associations were followed
2 and 3) are Earth’s largest and most intensely miner- by coeval 2.72 to 2.70 Ga deep-marine mantle-plume
alized Late Archean greenstone belts. Both terranes derived tholeiite–komatiite and tholeiite–calc-alkaline
have tectonic histories that are tightly constrained by arc associations [29]. The period 2.70 to 2.66 Ga saw
high-precision geochronology. The evolution of the episodic deep-water volcaniclastic sedimentation
Superior Province involved multiple komatiite, tho- (with black shales and layered mafic–ultramafic
leiite and calk-alkaline volcanic assemblages from sills) in the Kalgoorlie Terrane and regional emplace-
~2.78 Ga [43]. Submarine magmatism culminated ment of granitoids. This was followed by terrane
with the eruption of extensive suites of mantle accretion, orogeny and stabilization of continental
plume derived komatiites at 2.72 to 2.70 Ga. Exten- lithosphere between 2.65 and 2.62 Ga. Most other
sive hydrothermal activity, produced volcanic massive Late Archean greenstone terranes show similar his-
sulfide mineralization and BIF deposition in anoxic tories that parallel those recorded by marginal basins
arc-related basins. Arc and plume magmatism were to the Pacific (such as the Rocas Verdes Basin) during
followed by orogenic deformation, granitoid emplace- the Mesozoic breakup of Pangea [29]. The postulated
ment (by 2.68 Ga), stabilization of continental litho- period of plume breakout between 2.72 and 2.66 Ga
sphere and collision with the other cratons to form the [29,31] is recorded by elevated magmatism and high
Kenorland continent [16,18]. sea levels on all Late Archean cratons.

Fig. 3. A. Schematic diagrams (not to scale) illustrating changing environmental conditions and the tectonic evolution of the Pilbara Craton as a
type example of the rock record of the margin to an internal ocean during a full Late Archean to Paleoproterozoic global tectonic cycle. B.
Schematic diagrams (not to scale) illustrating changing environmental conditions and the tectonic evolution of representative convergent
margins during the Late Archean to Paleoproterozoic global tectonic cycle.
 M.E. Barley et al. / Earth and Planetary Science Letters 238 (2005) 156–171 163

Fig. 3 (continued).

The formation of Kenorland [16,18] and possible orogenic cycle, that also affected equivalent rocks in
collision of the Zimbabwe and Kaapvaal cratons at east Antarctica [46]. Central India and possibly east-
~2.6 Ga [44] provides evidence that Late Archean ern North China have similar histories from ~2.6 Ga
cratons started to aggregate into larger continents at culminating with orogeny and high-grade granulite
that time. Importantly granitoid–greenstone terranes facies metamorphism between ~2.5 and 2.42 Ga
and high-grade gneiss belts in the Gawler Craton, [47,48] corresponding to the aggregation and stabili-
Antarctica, India, and China [45–48] provide evidence zation of Indian cratons within a larger continent.
for a second cycle of convergent margin tectonics and Anticlockwise P–T-time paths in granulites are char-
collision of cratons between 2.6 and ~2.42 Ga. The acteristic of thickened and magmatically heated crust
Gawler Craton (Figs 2 and 3) contains 2.56 to 2.5 Ga in continental margin magmatic arcs such as the Cre-
ultramafic to felsic volcanic rocks (including ~2.51 taceous granulites of Fiordland in SW New Zealand
Ga plume-derived komatiites), metasedimentary [50].
rocks, and granitoids with compositions that are typi- The Pilbara and Kaapvaal are the only cratons with
cal of Archean granitoid–greenstone terranes inter- relatively complete and well-dated 2.6 to 2.4 Ga
preted to have formed at convergent continental supracrustal rock records (Figs. 2 and 3). Flexural
margins [49]. These were deformed, intruded by gran- reactivation of the southern Pilbara margin from
itoids and metamorphosed to high-grade (up to gran- ~2.59 Ga, following deposition of platform carbo-
ulite facies) during the 2.48 to 2.42 Ga Sleafordian nates, black shales and banded iron formations on a
164 M.E. Barley et al. / Earth and Planetary Science Letters 238 (2005) 156–171

passive margin (Carawine Dolomite, Roy Hill Shale, Hearst Dykes are dated at 2.47 and 2.45 Ga coeval
Marra Mamba Iron Formation, making up the Marra with bimodal volcanism at the base of the Huronian
Mamba Supersequence), led to deposition of a deep- Supergroup [56]. Continental extension and mafic
water carbonate ramp and then a condensed sequence large igneous provinces are also dated at 2.5 to 2.44
of black shales (respectively Paraburdoo and Bee Ga in Fennoscandia [57]. These igneous provinces are
Gorge Supersequences in the Wittenoom Dolomite). generally viewed as incipient rifting of the Kenorland
These are overlain by repeated cycles of deep-water continent [18], but the global picture with magmatism
shale, carbonate and banded iron formation of the coeval with BIF deposition on the continental margins
2.50–2.45 Ga Brockman Supersequence, with maxi- to the Pilbara and Kaapvaal, and with granitoid–
mum BIF deposition between 2.47–2.45 Ga coinci- greenstone terranes and high-grade gneiss belts in
dent with emplacement of a bimodal large igneous other cratons followed by low sea level and global
province [41,51,52]. The Brockman Supersequence is tectonic quiescence, is consistent with the ~2.5 to 2.45
the youngest well-dated unit that contains sulfides Ga event representing plume breakout during the
with D33S values N0.4x [1, 4], indicating that atmo- advanced stages of supercontinent assembly with lim-
spheric oxygen levels were still low at this time. The ited production of new continental crust, similar to the
Boolgeeda Iron Formation at the top of the Brockman Permo–Carboniferous event. The best age constraint
Supersequence is overlain by the upwards-shallowing on the final stages of Late Archean continental assem-
Turee Creek Supersequence comprising basin-plain bly is the 2.42 Ga end to the Sleafordian orogenic
(including resedimented glacial diamictites of the cycle [46] that coincides with foreland basin sedimen-
Meteorite Bore Member), carbonate-platform and flu- tation in the Pilbara and Kaapvaal cratons and the start
viodeltaic facies, deposited in a foreland basin [25], of a period with little evidence of orogeny (tectonic
indicating collision with another continental fragment, quiescence).
lower relative sea levels and a colder climate. Our Thus, the growing database of high precision geo-
ongoing work indicates that the youngest relatively chronology from most Archean cratons reinforces the
abundant, distinct population of detrital zircons in interpretation of Blake and Barley [39] that the
quartzites in the upper part of the Turee Creek Hamersley Province (Mount Bruce Supergroup) of
Group (overlying the Meteorite Bore Member) is the Pilbara Craton is the rock record of a Late Archean
2420 F 10 Ma, consistent with deposition of the to Paleoproterozoic 2.78 to 2.40 Ga global tectonic
group between 2.45 and 2.40 Ga. cycle involving continental breakup and the possible
The Kaapvaal succession comprises 2.6–2.5 Ga formation of a supercontinent. Plume breakout during
upward-deepening carbonates overlain by BIF, includ- the breakup phase of this cycle resulted in enhanced
ing shallow-water reduced facies [53], and an upper- submarine and convergent margin magmatism corre-
most predominantly clastic sedimentary succession sponding to one of the most prodigious periods of
recording conversion to compressional tectonics continental crust formation in Earth history
[54]. Zircons in tuff layers in the BIF indicate that [20,29,31]. These continental fragments appear to
volcanism and BIF deposition are the same age as in have amalgamated to form Earth’s first supercontinent
the Pilbara [55]. The 2.59–2.40 Ga tectonic histories by ~2.40 Ga.
of the Pilbara and Kaapvaal cratons are interpreted to
reflect the conversion from trailing passive margins of
an internal ocean to convergent continental margins of 4. Paleoproterozoic tectonics — 2.40 to ~2.2 Ga
a Tethyan-style external ocean [23].
Barley et al. [51] and Kump et al. [7] suggest that Widespread continental erosion and the scarcity of
the period 2.5–2.45 Ga is a global plume breakout easily datable supracrustal rocks, are consistent with
event recorded by mafic magmatism on several con- the existence of a 2.4 Ga high-standing superconti-
tinents as well as high sea-levels and BIF deposition. nent, but hamper interpretation of Earth history in the
Large igneous provinces between 2.5–2.45 Ga also critical period between 2.40 and 2.2 Ga. Most cratons
occur in the Superior, Wyoming and Fennoscandian preserve evidence of intracontinental extension and
Cratons. In the Superior Craton, the Matachewan and episodic mafic magmatism from ~2.4 to 2.2 Ga,
 M.E. Barley et al. / Earth and Planetary Science Letters 238 (2005) 156–171 165

with continental, rift and passive margin environments Hearst Dyke swarms) overlain by continental sedi-
variably preserved in the Kaapvaal, Superior, Hearne, mentary rocks that are in turn overlain by three cycles
Rae, Wyoming and Fennoscandian cratons. In the that begin with glaciogene sedimentary rocks and pass
Pilbara, an ~380 Ma hiatus separates the Turee upwards into basinal mudrocks and carbonates and
Creek Supersequence from oxidized clastic sedimen- then to fluvial or shallow-marine sandstones deposited
tary rocks, basalts and carbonates of the Lower Wyloo under more temperate conditions. Successions below
Group that were deposited at ~2.03 Ga [58]. and above the first glacial diamictite (the Ramsay
The Kaapvaal Craton (Fig. 2) has an important and Lake Formation) contain detrital uraninite and pyrite,
poorly dated succession of continental volcanic and although shales immediately underlying and overlying
sedimentary rocks deposited between 2.4 and 2.0 Ga the Ramsay Lake Formation contain the oldest sedi-
[59]. In the Transvaal Basin, the Lower Pretoria mentary sulfides with D33S values b0.4x suggesting
Group unconformably overlying N 2.40 Ga BIFs, car- that levels of atmospheric oxygen were N10 5 PAL at
bonates and clastic sedimentary rocks comprises the this time [5]. Carbonates of the Espanola Formation
Rooithoogte–Duitschland and Timeball Hill forma- sitting on the second glacial diamictite (the Bruce
tions was deposited in a continental basin open to Formation) have negative carbon isotope values and
an ocean to the southwest. The Rooithoogte Forma- are likely correlative with carbonates of the Duitsch-
tion comprises a lower glacial diamictite, overlain by land Formation in South Africa [59,65]. The third
basalt, sandstones and black shales containing the Huronian glacial diamictite occurs in the lower part
oldest well-dated syngenetic pyrite (2.32 Ga by Re- of the Gowganda Formation, the lowermost unit of the
Os isochron [60]) with D33S values less than 0.4x upper Huronian. The upper part of the Gowganda
indicating that atmospheric oxygen was N10 5 PAL Formation contains redbeds, with an oxidized paleo-
[3]. The correlative Duitschland Formation also con- sol developed below the overlying Lorrain Formation
tains glacial diamictites overlain by cap carbonates [66]. The upper Huronian is interpreted as comprising
with negative carbon isotope values, sandstones mark- the transition from a rifted to passive continental
ing a prominent sequence boundary, and overlying margin [19]. The mafic Nipissing sills dated at
carbonates with carbon isotope values as high as 2217 F 1.6 Ma provide the minimum age for deposi-
+ 10x [59]. Sandstones overlying the Lower Timeball tion [67]. Thin carbonates in the upper Huronian
Hill shales contain extensive shallow-water deposits Supergroup, above the Gowganda Formation diamic-
of hematite oolites and pisolites indicating oxygenated tite have highly positive carbon isotope values similar
conditions. These are in turn overlain by younger to those found worldwide in carbonates with ages
diamictites, the oldest partially oxidized lateritic between N 2.22 and 2.1 Ga [65,68]. The Wyoming
weathering profile [61] developed on continental and Fennoscandian cratons also contain a similar
lavas of the Hekpoort Formation. This succession is record of continental sedimentation, rifting, glacia-
generally correlated with altered submarine lavas of tion, carbon isotope excursions and subsequent
the ~2.22 Ga Ongeluk andesite that overlie the Mak- breakup [65,69,70].
ganyene diamictites [62,63] and are conformably Bekker et al. [59] have suggested that it is possible
overlain by BIF and manganese deposits of the Hota- to correlate early Paleoproterozoic successions based
zel Formation and the Mooidrai dolomite indicating on glacial deposits and carbonate chemostratigraphy
rifting and flooding of the continental margin. In both inferring that the second and third Huronian diamic-
the Kaapvaal and Pilbara cratons economically impor- tites are equivalent to diamictites at the base of the
tant BIF-hosted hematite deposits formed between 2.2 Rooithoogte–Duitschland Formation and top of the
and 2.0 Ga [58,62,64]. Timeball Hill Formation, respectively. If so the oldest
In the Superior Craton a major hiatus separates the Huronian diamictite and the N2.40 Ga Turee Creek
greenstone and granite–gneiss terrane from the over- diamictite may also be equivalents, as both occur in
lying Huronian Supergroup (Fig. 2). The Huronian conformable successions above 2.45 Ga volcanic
Supergroup comprises four tectonostratigraphic cycles rocks, with the first widespread glaciation following
[19]. The oldest contains 2.48 to 2.45 Ga bimodal relatively soon (perhaps within 10 to 40 million
volcanic rocks (coeval with the Matachewan and years) after the 2.5–2.45 Ga plume breakout event.
166 M.E. Barley et al. / Earth and Planetary Science Letters 238 (2005) 156–171

The first and second glacials are further constrained computer models which predict that this event
by the ~2.32 Ga age of the Rooithoogte Formation involved orders of magnitude greater heat flux and
[60], with a minimum age for the third provided by mafic magmatism than Phanerozoic plume breakout
the 2.22 Ga age of the Ongeluk lavas [63] and events [12]. Granitoid–greenstone and granite–gneiss
Nipissing sills [67] Although there is not yet suffi- terranes started to amalgamate to form stable cratons
cient precise geochronology to allow chronostrati- and continents from ~2.65 Ga (e.g. Kenorland). How-
graphic correlation of 2.4 to 2.2 Ga tectonic events ever, shallow-marine carbonate sedimentation after
between cratons, the picture that emerges is one of 2.6 Ga was followed by deposition of black shales
widespread continental erosion (low relative sea and BIF on the trailing continental margins of the
level) and sedimentation in intracontinental basins Pilbara and Kaapvaal cratons coincident with plume
followed by rifting, flooding of continental margins breakout from ~2.50 to 2.45 Ga. The following return
and breakup of the supercontinent, with some rifted to lower relative sea levels and orogenic sedimenta-
continental margins formed by ~2.32 Ga. tion after 2.45 Ga reflects possible terrane collision
and conversion to the margin of an external ocean.
Importantly, granitoid–greenstone terranes, and gneiss
5. Global tectonics environmental change and the belts in Australia, Antarctica, India and North China
rise of atmospheric oxygen record histories of convergent margin tectonics, mag-
matism, orogeny and high-grade metamorphism
This review suggests that although containing the between 2.6 and ~2.42 Ga culminating in the amalga-
rock record of the largest episode of plume breakout mation of cratons into a larger continent or super-
recorded in Earth history, the tectonostratigraphic continent (e.g. amalgamation and stabilization of
records of Late Archean and Early Paleoproterozoic Indian cratons, [47]).
terranes record a global tectonic cycle involving con- Archean continental sediments contain detrital pyr-
tinental breakup followed by growth and aggregation ite, uraninite and siderite [71,72]. Highly carbonac-
of continental fragments to form large continents or a eous shales from this period are also not enriched in
supercontinent that was broadly similar in style, dura- redox sensitive elements [73], early diagenetic pyrite
tion and relative timing to Phanerozoic tectonic cycles. has d 34 S compositions consistent with low seawater
Were relative sea level, climate and conditions in the sulfate concentrations [74], and sedimentary sulfide
atmosphere and hydrosphere also linked to tectonics? NMDF signals N0.4 per mil D33S [1,2,4]. All these
Late Archean (Figs. 2 and 3) continental breakup features are consistent with an anoxic atmosphere and
started at ~2.78 Ga with stable, high-standing (low hydrosphere, although oxygenic photosynthesis dates
relative sea level) continental crust in the Pilbara and from at least 2.71 Ga [11] and may be older. In this
Kaapvaal Cratons (possible Vaalbra continent). Sub- regard it is noteworthy that black shales from the
sequent intracontinental extension, sedimentation, period 2.72–2.65 Ga have some of the most negative
volcanism and rifting were succeeded by flooding of d 13C isotopic compositions of kerogen and total
continental margins and deposition of sulphidic black organic carbon recorded, most likely reflecting
shales during a 2.71 to 2.65 Ga period of plume methane production and recycling by methanogens
breakout. These passive margins to an Atlantic-type and methanotrophs [75,76].
internal ocean saw a return to lower relative sea levels If the Turee Creek and oldest Huronian glaciation
by 2.59 Ga. Coeval 2.78 to 2.6 Ga granitoid–green- represent the Earth’s oldest widespread glaciation, the
stone terranes record growth and stabilization of juve- rise of atmospheric O2 to N 10 5 PAL occurred at some
nile continental crust via the opening and closure of time between 2.47 and 2.40 Ga, following a period of
marine basins (including BIF deposition), terrane plume breakout and coinciding with assembly of the
accretion and magmatism at convergent continental Earth’s first supercontinent. If oxygenic photosynthe-
margins of the external ocean. The period of plume sizers had evolved by 2.71 Ga, why did reducing
breakout between 2.72 and 2.65 was the most prodi- conditions persist for more than 250 million years?
gious episode of juvenile continental crust formation The most likely explanation for this is that atmospheric
preserved in Earth history [20,29,31] consistent with O2 levels were not able to rise until photosynthetic
 M.E. Barley et al. / Earth and Planetary Science Letters 238 (2005) 156–171 167

oxygen production was able to satisfy the available vide favorable conditions for this change (Fig. 4). The
global oxygen sinks. The magnitude of the oxygen major flux of reduced gasses, hydrothermal alteration
sinks (reduced volcanic gasses, plus new ocean and and serpentinization of oceanic crust, and oxidation of
continental crust and BIF) produced during the 2.71 to ferrous iron to form extensive BIFs would have been
2.66 Ga plume breakout, compared to younger plume major global O2 sinks during 2.51–2.45 Ga plume
breakout events, would have delayed oxidation of the breakout. High-grade metamorphism during 2.5 to
oceans and atmosphere, favoring anerobic biogenic 2.42 Ga continental assembly may also have been
production of CH4 (e.g. [76]), enhanced atmospheric an important oxygen sink producing a significant
CO2 and CH4 levels, and the greenhouse conditions flux of reduced gasses, although coeval weathering
necessary to maintain temperate conditions. The of uplifted orogens would draw down atmospheric
increased hydrothermal flux of iron and deposition CO2. Thus the decreased flux of reduced gasses
of iron oxides as BIF during plume breakout would after 2.45 Ga coupled with a possible increase in
have also limited phosphorous abundance and inhib- ocean productivity and the filling of oxygen sinks
ited ocean organic productivity and rates of photo- produced during plume breakout and continental
synthesis contributing to low concentrations of assembly may have resulted in the global flux of
atmospheric oxygen [9,10]. Reduced conditions with reduced gasses falling below that of photosynthetic
atmospheric O2 b10 5 PAL continued during the 2.51 oxygen production leading to a rise in atmospheric O2
to 2.45 Ga plume breakout event. to N 10 5 PAL which was accompanied by loss of the
The Kump et al. model [7] for irreversible oxidation CH4-rich greenhouse atmosphere resulting in the
of the atmosphere due to an increase in the oxidation Earth’s first widespread glaciation. The persistence
state of volcanic gasses between the two plume break- of detrital pyrite and uraninite in 2.45 to 2.40 Ga
out events accompanied by filling of crustal O2 sinks is sediments suggests that deep weathering oxidative
the only model that specifically predicts rapid initial weathering of continental crust in terrestrial environ-
oxidation of the atmosphere (sufficient to reduce the ments had not yet occurred.
sulfur NMDF signal) during or before the 2.51 to 2.45 A plausible climatic scenario is rapid reduction of
Ga plume breakout. The main objection raised to this atmospheric CH4 levels due to rising atmospheric O2
hypothesis is that the abundances of redox sensitive following 2.51 to 2.45 Ga plume breakout during
trace elements indicate little difference between the supercontinent assembly accompanied by draw
redox state of the sources of Archean and modern down of CO2 resulting in the first widespread glacia-
Mg-rich mafic magmas [77]. However, most preserved tion (Meteorite Bore Member of the Turee Creek
Archean Mg-rich volcanic rocks are thought to have Group, Ramsay Lake Formation of the Huronian
formed in either mantle-plume or subduction-related Supergroup). This was possibly followed by a period
environments [29,43,78] and therefore may reflect of rising atmospheric CO2 from ongoing mafic vol-
subducted oceanic crust and not be representative of canism, and inhibited weathering during tectonic
the oxidation state of magmas derived from the upper quiescence with a return to milder climates. Atmo-
mantle at Late Archean oceanic spreading centres. spheric O2, CH4 and CO2 levels may have fluctuated
Thus changing oxidation states of volcanic gasses between glacial and interglacial periods [65,80]. Deep
[7,8,12] remains a possible explanation for the initial weathering during the initial stages of rifting and
rise of atmospheric oxygen. If global deposition of breakup reduced atmospheric CO2 and resulted in
BIFs declined at the end of the 2.51 to 2.45 Ga further glaciations at ~2.32 Ga. with a positive carbon
plume breakout [51,79] ocean productivity and the isotope excursion between the second and third sug-
rate of photosynthesis may also have increased at gesting increased organic carbon burial and a further
this time contributing to rising O2 levels [9,10]. rise in atmospheric oxygen [59] coinciding with the
If rapid loss of hydrogen to space from a CH4-rich oldest evidence of extensive oxidative surface weath-
atmosphere coupled with filling of the main global ering. If the oxidation state of volcanic gasses has not
sinks for oxygen as argued by Catling et al. and changed significantly since the Archean, lower mag-
Kasting [6,13] provides the explanation for the rise nitude post-Archean plume breakout events (com-
of oxygen, Early Paleoproterozoic tectonics may pro- pared to the 2.72 to 2.65 Ga event) clearly have not
168 M.E. Barley et al. / Earth and Planetary Science Letters 238 (2005) 156–171

Fig. 4. Schematic diagrams (not to scale) illustrating the links between tectonics and changing global environmental conditions between 2.5 and
2.22 Ga.
 M.E. Barley et al. / Earth and Planetary Science Letters 238 (2005) 156–171 169

produced sufficient volumes of reduced gasses and [2] J. Farquhar, B.A. Wing, Multiple sulfur isotopes and the
crustal oxygen sinks to reverse oxidation of the atmo- evolution of the atmosphere, Earth Planet. Sci. Lett. 213
(2003) 1 – 13.
sphere, although widespread deposition of BIF during [3] A. Bekker, H.D. Holland, P.L. Wang, D. Rumble, H.J. Stein,
a major period of increased growth of juvenile con- J.L. Hannah, L.L. Coetzee, N.J. Beukes, Dating the rise of
tinental crust and plume breakout between 2.0 and 1.8 atmospheric oxygen, Nature 427 (2004) 117 – 120.
Ga [20,79] suggests a return to anoxic conditions in [4] S.J. Mojzsis, C.D. Coath, J.P. Greenwood, K.D. McKeegan,
T.M. Harrison, Mass-independent isotope effects in Archean
the ocean at that time [10].
(2.5 to 3.8 Ga) sedimentary sulfides determined by ion
Although there remains uncertainty over its exact microprobe analysis, Geochim. Cosmochim. Acta 67 (2003)
timing and cause, this analysis indicates that irrever- 1635 – 1658.
sible oxidation of the atmosphere most likely occurred [5] B.A. Wing, A. Bekker, J. Farquhar, A.J. Kaufman, D. Rum-
during or soon after the 2.51 to 2.45 Ga plume break- ble, Atmospheric chemistry of SO2 during the earliest Hur-
out coinciding with a decline in BIF deposition and onian glaciation, Geol. Soc. Am. Abstracts Programs 36 (5)
(2004) 339.
maximum assembly of the Earth’s first supercontinent. [6] D.C. Catling, K.J. Zahnle, C.P. McKay, Biogenic methane,
Systematic S and Fe isotope studies linked to geo- hydrogen escape, and irreversible oxidation of early Earth,
chronology of diagenetic sulfides and phosphates in Science 293 (2001) 839 – 843.
the upper Brockman and Turee Creek supersequences [7] L.R. Kump, J.F. Kasting, M.E. Barley, Rise of atmospheric
and Lower Huronian Supergroup, and the geochronol- oxygen and the upside down Archean mantle, Geochem.
Geophys. Geosyst. 2 (2001) 2000GC000114.
ogy of low temperature alteration in Late Archean [8] H.D. Holland, Volcanic gasses, black smokers and the great
sedimentary basins have the potential to unambigu- oxidation event, Geochim. Cosmochim. Acta 66 (2002)
ously resolve the timing of specific events related to 3811 – 3826.
the rise of atmospheric oxygen. It is clear that better [9] C.J. Bjerrum, D.E. Canfield, Ocean productivity before about
chronostratigraphic resolution of the geological and 1.9 Gyr ago limited by phosphorous adsorption onto iron
oxides, Nature 417 (2002) 159 – 162.
geochemical records will ultimately resolve the true [10] D.E. Canfield, The early history of atmospheric oxygen, Annu.
complexity of coevolving Earth systems during the Rev. Planet. Sci. 33 (2005) 1 – 36.
Late Archean and Early Paleoproterozoic and may [11] J.J. Brocks, G.A. Logan, R. Buick, R.E. Summons, Archean
ultimately allow a quantitative view of the dynamic molecular fossils and the early rise of eukaryotes, Science 285
interplay between the source and sinks for oxygen. (1999) 1033 – 1036.
[12] J.F. Kasting, D.H. Eggler, S.P. Raeburn, Mantle redox evolu-
tion and the oxidation state of the atmosphere, J. Geol. 101
(1993) 245 – 257.
Acknowledgements [13] J.F. Kasting, The rise of atmospheric oxygen, Science 293
(2001) 819 – 820.
[14] G.F. Davies, Punctuated tectonic evolution of the Earth, Earth
We acknowledge support from the Australian
Planet. Sci. Lett. 136 (1995) 363 – 379.
Research Council, the Minerals and Energy Research [15] R.D. Nance, T.R. Worsley, J.B. Moody, Post-Archean biogeo-
Institute of Western Australia, NASA Astrobiology chemical cycles and long-term episodicity in tectonic pro-
and the Carnegie Institution of Washington. Lee cesses, Geology 14 (1986) 514 – 518.
Kump is thanked for prompting this exploration of [16] H. Williams, P.F. Hoffman, J.F. Lewry, J.W.H. Monger, T.
dynamic links between Archean–Paleoproterozoic Rivers, Anatomy of North America: thematic geologic
portrayals of the continent, Tectonophysics 187 (1991)
global tectonics and environmental conditions, for 117 – 134.
reading early drafts of the manuscript and discussion [17] E.S. Cheney, Sequence stratigraphy and plate tectonic signifi-
and comments that significantly improved them. We cance of the Transvaal succession in southern Africa and its
appreciate helpful reviews by David Evans, Jim Kast- equivalent in western Australia, Precambrian Res. 79 (1996)
ing and an anonymous reviewer. 3 – 24.
[18] L.B. Aspler, J.R. Chiarenzelli, Two Neoarchean superconti-
nents? Evidence from the Paleoproterozoic, Sediment. Geol.
120 (1998) 75 – 104.
References [19] G.M. Young, D.G.F. Long, C.M. Fedo, H.W. Nesbitt, Paleo-
proterozoic Huronian basin: product of a Wilson cycle punc-
[1] J. Farquhar, H.M. Bao, M. Thiemens, Atmospheric influence tuated by glaciations and a meteorite impact, Sediment. Geol.
of Earth’s earliest sulfur cycle, Science 289 (2000) 756 – 758. 141–142 (2001) 233 – 254.
170 M.E. Barley et al. / Earth and Planetary Science Letters 238 (2005) 156–171

[20] K.C. Condie, Mantle Plumes and Their Record in Earth [39] T.S. Blake, M.E. Barley, Tectonic evolution of the Late
History, Cambridge University Press, New York, NY, 2001, Archaean to Early Proterozoic Mount Bruce Megasequence,
306 pp. Tectonics 11 (1992) 1415 – 1425.
[21] J.B. Murphy, R.D. Nance, Supercontinent model for the con- [40] T.S. Blake, R. Buick, S.J.A. Brown, M.E. Barley, Geochro-
trasting character of Late Proterozoic orogenic belts, Geology nology of a late Archaean flood basalt province in the Pilbara
19 (1991) 469 – 472. Craton, Australia: constraints on basin evolution, volcanic and
[22] J.B. Murphy, R.D. Nance, Do supercontinents introvert or sedimentary accumulation, and continental drift rates, Precam-
extrovert?: Sm–Nd isotope evidence, Geology 31 (2003) brian Res. 133 (2004) 143 – 173.
873 – 876. [41] B. Krapez, M.E. Barley, A.L. Pickard, Hydrothermal and
[23] B. Krapez, Sequence-stratigraphic concepts applied to the resedimented origins for the precursor sediments to banded
identification of depositional basins and global tectonic cycles, iron formations: sedimentological evidence from the early
Aust. J. Earth Sci. 44 (1997) 1 – 36. Paleoproterozoic Brockman Supersequence of Western Aus-
[24] P.F. Hoffman, Did the breakout of Laurentia turn Gondwana tralia, Sedimentology 50 (2003) 979 – 1011.
inside out? Science 252 (1991) 1409 – 1412. [42] D.R. Nelson, A.F. Trendall, W. Altermann, Chronological
[25] B. Krapez, Sequence-stratigraphic concepts applied to the correlations between the Pilbara and Kaapvaal Cratons, Pre-
identification of basin-filling rhythms in Precambrian succes- cambrian Res. 97 (1999) 165 – 189.
sions, Aust. J. Earth Sci. 43 (1996) 355 – 380. [43] D.A. Wyman, R. Kerrich, A. Polat, Assembly of an Archean
[26] B.C. Storey, The role of mantle plumes in continental cratonic mantle lithosphere and crust: plume–arc interaction in
breakup: case histories from Gondwanaland, Nature 377 the Abitibi–Wawa subduction–accretion complex, Precam-
(1995) 301 – 308. brian Res. 115 (2002) 37 – 62.
[27] H.C. Larson, Latest pulse of Earth: evidence for a mid-Cretac- [44] B.S. Kamber, G.G. Biino, J.R. Wijbrans, G.R. Davies, I.M.
eous superplume, Geology 19 (1991) 547 – 550. Villa, Archean granulites of the Limpopo Belt, Zimbabwe: one
[28] H.C. Larson, Geological consequences of superplumes, Geol- slow exhumation or two rapid events, Tectonics 15 (1996)
ogy 19 (1991) 963 – 966. 1414 – 1430.
[29] M.E. Barley, B. Krapez, D.I. Groves, R. Kerrich, The Late [45] S.J. Daly, C.M. Fanning, M.C. Fairclough, Tectonic evolution
Archaean bonanza: metallogenic and environmental conse- and exploration potential of the Gawler Craton, South Aus-
quences of the interaction between mantle plumes, lithospheric tralia, AGSO, J. Aust. Geol. Geophys. 17 (1998) 145 – 168.
tectonics and global cyclicity, Precambrian Res. 91 (1998) [46] C.M. Fanning, D.H. Moore, V.C. Bennet, S.J. Daly, The
65 – 90. bMawson ContinentQ Archaean to Proterozoic crust in the
[30] T.R. Worsley, R.D. Nance, J.B. Moody, Tectonic cycles and East Antarctic Shield and Gawler Craton, Australia, Geol.
the history of the Earth’s biogeochemical record, Paleoocea- Soc. Aust. Abstracts 41 (1996) 135.
nography 1 (1986) 4111 – 4118. [47] H.J. Stein, J.L. Hannah, A. Zimmerman, R.J. Markey, S.C.
[31] K.C. Condie, Episodic continental growth and superconti- Sarkar, A.B. Pal, A 2.5 Ga porphyry Cu–Mo–Au deposit at
nents: a mantle avalanche connection? Earth Planet. Sci. Malanjkhand, central India: implications for Late Archean
Lett. 163 (1998) 97 – 108. continental assembly, Precambrian Res. 134 (2004) 189 – 226.
[32] N. Eyles, G.M. Young, Geodynamic controls on glaciation in [48] G. Zhao, M. Sun, S.A. Wilde, Correlations between the East-
Earth history, in: M. Deynoux, J.G.M. Miller, E.W. Domacket ern Block of the North China Craton and the South Indian
(Eds.), Earth’s Glacial Record, Cambridge University Press, Block of the Indian Shield: an Archaean to Palaeoproterozoic
Cambridge, 1994, pp. 1 – 28. link, Precambrian Res. 122 (2003) 201 – 233.
[33] J.C. Crowell, Pre-Mesozoic ice ages: their bearing on under- [49] G. Swain, A. Woodhouse, M. Hand, K. Barovich, M.
standing the climate system, Geol. Soc. Am., Mem., vol. 192, Schwartz, Late Archaean tectonothermal history of the proto
1999, 105 pp. Gawler Craton: Geochemical and isotopic constraints, Precam-
[34] P.F. Hoffman, A.J. Kaufman, G.P. Halverson, D.P. Schrag, brian Res. (in press).
A Neoproterozoic snowball Earth, Science 281 (1998) [50] J.Y. Bradshaw, Origin and metamorphic history of an early
1342 – 1346. Cretaceous polybaric granulite terrain, Fiordland southwest
[35] J.M. Franklin, J.W. Lydon, D.F. Sangster, Volcanic associated New Zealand, Contrib. Mineral. Petrol. 103 (1989) 346 – 360.
massive sulfide deposits, Econ. Geol. 75th Anniv. Vol. (1981) [51] M.E. Barley, A.L. Pickard, P.J. Sylvester, Emplacement of a
485 – 627. large igneous province as a possible cause of banded iron
[36] F.B. Van Houten, M.A. Arthur, Temporal patterns among formation 2.45 billion years ago, Nature 385 (1997) 55 – 58.
Phanerozoic oolitic ironstones and ocean anoxia, Geol. Soc. [52] A.L. Pickard, SHRIMP U–Pb zircon ages of tuffaceous
London, Spec. Publ. 46 (1989) 33 – 49. mudrocks in the Brockman Iron Formation of the Hamers-
[37] A. Ishiwatari, Circum-Pacific multiple ophiolite belts, Proc ley Range, Western Australia, Aust. J. Earth Sci. 49 (2002)
29th International Geol. Congr. Part D, 1994, pp. 7 – 28. 491 – 507.
[38] W.D. Huff, S.M. Bergstrom, D.R. Kolata, Gigantic Ordovician [53] N.J. Beukes, Palaeoenvironmental setting of iron formation in
volcanic ash fall in North America and Europe: biological, the depositional basin of the Transvaal Supergroup, South
tectonomagmatic and event-stratigraphic significance, Geol- Africa, in: A.F. Trendall, R.C. Morris (Eds.), Iron Formations:
ogy 20 (1992) 875 – 878. Facts and Problems, Elsevier, Amsterdam, 1983, pp. 69 – 129.
 M.E. Barley et al. / Earth and Planetary Science Letters 238 (2005) 156–171 171

[54] W. Altermann, I.W. Halbich, Structural history of the south- [67] A.J. Andrews, A. Masliwec, W.A. Morris, L. Owsiacki, D.
western corner of the Kaapvaal Craton and the adjacent Nama- York, The silver deposits at cobalt and Gowganda Ontario II:
qua realm: new observations and reappraisal, Precambrian an experiment in age determinations employing radiometric
Res. 52 (1991) 133 – 166. and paleomagnetic measurements, Can. J. Earth Sci. 23 (1986)
[55] A.L. Pickard, SHRIMP U–Pb zircon ages for the Paleoproter- 1507 – 1518.
ozoic Kuruman Iron Formation, Northern Cape Province, [68] J.A. Karhu, H.D. Holland, Carbon isotopes and the rise of
South Africa: evidence for simultaneous BIF deposition on atmospheric oxygen, Geology 24 (1996) 867 – 870.
Kaapvaal and Pilbara Cratons, Precambrian Res. 125 (2003) [69] A. Bekker, J.A. Karhu, K.A. Eriksson, A.J. Kaufman, Che-
355 – 380. mostratigraphy of Paleoproterozoic carbonate successions of
[56] L.M. Heaman, Global mafic magmatism at 2.45 Ga: remnants the Wyoming Craton: tectonic forcing of biogeochemical
of an ancient large igneous province? Geology 25 (1997) change? Precambrian Res. 120 (2003) 279 – 325.
299 – 302. [70] V.A. Melezhik, B.A. Sturt, General geology and evolution-
[57] Y.V. Amelin, L.M. Heaman, V.S. Semenov, U–Pb geochronol- ary history of the early Proterozoic Polmak–Pasvik–
ogy of layered mafic intrusions in the eastern Baltic Shield: Pechenga–Immandra/Varzuga–Ust’ Ponoy Greenstone belt
implications for the timing and duration of Paleoproterozoic in the northeastern Baltic Shield, Earth Sci. Rev. 36 (1994)
continental rifting, Precambrian Res. 75 (1995) 31 – 46. 205 – 241.
[58] S. Mueller, I.R. Fletcher, B. Krapez, M.E. Barley, A post 2.01 [71] S.M. Roscoe, Huronian rocks and uraniferous conglomerates
Ga anorogenic origin for giant hematite ore bodies in the in the Canadian Shield, Geol. Surv. Can. Pap. 68–40 (1969)
Hamersley Province, Western Australia, Geology 33 (2005) (205 pp.).
577 – 580. [72] B. Rasmussen, R. Buick, Redox state of the Archean atmo-
[59] A. Bekker, A.J. Kaufman, J.A. Karhu, N.J. Beukes, Q.D. sphere: evidence from detrital heavy minerals in 3250–2750
Swart, L.L. Coetzee, K.A. Eriksson, Chemostratigraphy of Ma sandstones from the Pilbara Craton, Australia, Geology 27
the Paleoproterozoic Duitschland Formation South Africa: (1999) 115 – 118.
implications for coupled climate change and carbon cycling, [73] W. Yang, H.D. Holland, The redox-sensitive trace elements,
Am. J. Sci. 301 (2001) 261 – 285. Mo, U, and Re in Precambrian carbonaceous shales: indicators
[60] J.L. Hannah, A. Bekker, H.J. Stein, R.J. Markey, H.D. Hol- of the great oxidation event, Geol. Soc. Am. Abstracts Pro-
land, Primitive Os and 2316 Ma age for Paleoproterozoic grams 34 (2002) 381.
glacial events and the rise of oxygen, Earth Planet. Sci. Lett. [74] K.S. Habicht, M. Gade, B. Thamdrup, P. Berg, D.E. Canfield,
225 (2004) 43 – 52. Calibration of sulfate levels in the Archean ocean, Science 298
[61] N.J. Beukes, H. Dorland, J. Gutzmer, M. Nedachi, H. Ohmoto, (2002) 2372 – 2374.
Tropical laterites, life on land, and the history of atmospheric [75] J.M. Hayes, I.R. Kaplan, K.W. Wedeking, Precambrian
oxygen in the Paleoproterozoic, Geology 30 (2002) 491 – 494. organic geochemistry, preservation of the record, in: J.W.
[62] D.A. Evans, N.J. Beukes, J.L. Kirschvink, Low-latitude glacia- Schopf (Ed.), The Earth’s Earliest Biosphere, Princeton Uni-
tion in the Paleoproterozoic era, Nature 386 (1997) 262 – 266. versity Press, Princeton N.J., 1983, pp. 93 – 134.
[63] D.H. Cornell, S.S. Schutte, B.L. Eglinton, The Ongeluk basal- [76] K.-U. Hinrichs, Microbial fixation of methane carbon at 2.7
tic andesite formation in the Griquatown West, South Africa: Ga: was an anaerobic mechanism possible? Geochem. Geo-
submarine alteration in a 2222 Ma Proterozoic sea, Precam- phys. Geosyst. 3–7 (2002) 1 – 10.
brian Res. 79 (1996) 101 – 124. [77] D. Canil, Vanadium partitioning and the oxidation state of
[64] D.A. Evans, N.J. Beukes, J.L. Kirschvink, Paleomagnetism of Archean komatiite magmas, Nature 389 (1997) 842 – 845.
a lateritic paleoweathering horizon and overlying Paleoproter- [78] T.L. Grove, S.W. Parman, Thermal evolution of the Earth as
ozoic red beds from south Africa: implications for the Kaap- recorded by komatiites, Earth Planet. Sci. Lett. 219 (2004)
vaal apparent polar wander path and a confirmation of 173 – 187.
atmospheric oxygen enrichment, J. Geophys. Res. 107 (B12) [79] A.E. Isley, D.H. Abbott, Plume-related mafic volcanism and
(2002) 2326, doi:10.1029/2001JB000432. the deposition of banded iron formation, J. Geophys. Res. 104
[65] A. Bekker, A.J. Kaufman, J.A. Karhu, K.A. Eriksson, Evi- (1999) 15461 – 15477.
dence for early Paleoproterozoic cap carbonates in North [80] A.A. Pavlov, A. Bekker, Rises and falls of atmospheric oxygen
America, Precambrian Res. 137 (2005) 167 – 206. in the early Paleoproterozoic, Geol. Soc. Am. Abstracts Pro-
[66] R.H. Rainbird, H.W. Nesbitt, J.A. Donaldson, Formation and grams 36-5 (2004) 256.
diagenesis of a sub-Huronian saprolith: comparison with a
modern weathering profile, J. Geol. 98 (1990) 801 – 822.